Hydrological and chemical characteristics of karst groundwater and carbon flux estimation

Abstract

In order to understand the hydrological changes of groundwater in karst areas, the Xiangxi River Basin in the western Hubei karst area was selected for water sample collection and monitoring. A carbon flux estimation model was established by combining support vector machine regression and sparrow search algorithm to achieve groundwater carbon flux estimation. The results show that the average range of dissolved oxygen in karst groundwater is 9.65~10.79mg · L−1, with an overall high content. The anions in groundwater are mainly HCO3−, and the cations are mainly Ca2+. The pH value of the water sample is mainly between 7.28 and 8.16, showing weak alkalinity. The mineralization value of the water sample is distributed between 81.36~323.41mg/L, and the groundwater hydrochemical types are mainly HCO3 Ca type and HCO3 Ca · Mg type. The carbon flux estimation model for karst groundwater has reduced the regression error value by 43.6% compared to the unimproved support vector machine. This indicates that improving the model can improve the accuracy of carbon flux estimation. By analyzing the hydrochemical characteristics, the resource characteristics, environmental effects, and ecological significance of karst groundwater can be understood, providing scientific basis for management and decision-making in related fields..

Keywords

• Karst groundwater
• hydrological and chemical characteristics
• support vector machine regression
• sparrow search algorithm
• carbon flux
• Eaux souterraines karstiques